diff options
| author | Cherry Zhang <cherryyz@google.com> | 2016-07-06 10:04:45 -0400 |
|---|---|---|
| committer | Cherry Zhang <cherryyz@google.com> | 2016-07-16 03:13:22 +0000 |
| commit | 7d70f84f547a1b60279985fa91c407ddfde9bd64 (patch) | |
| tree | afc916dc065ed729b4466c274437f01790083fce /src/runtime | |
| parent | 6adb97bde72b97310f9a75a4e286cd2ef236b271 (diff) | |
| download | go-7d70f84f547a1b60279985fa91c407ddfde9bd64.tar.xz | |
[dev.ssa] cmd/compile: add floating point optimizations in SSA for ARM
Add some simplification rules for floating point ops.
cmd/internal/obj/arm supports instructions that compare FP register
to 0, but runtime softfloat simulator does not. This CL adds these
instructions to softfloat simulator as well.
Updates #15365.
Change-Id: I29405b2bfcb4c8cf106cb7a1a811409fec91b170
Reviewed-on: https://go-review.googlesource.com/24790
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Diffstat (limited to 'src/runtime')
| -rw-r--r-- | src/runtime/softfloat_arm.go | 18 |
1 files changed, 18 insertions, 0 deletions
diff --git a/src/runtime/softfloat_arm.go b/src/runtime/softfloat_arm.go index 5f609c80d3..802a151fbf 100644 --- a/src/runtime/softfloat_arm.go +++ b/src/runtime/softfloat_arm.go @@ -464,6 +464,24 @@ execute: } return 1 + case 0xeeb50bc0: // D[regd] :: 0 (CMPD) + cmp, nan := fcmp64(fgetd(regd), 0) + m.fflag = fstatus(nan, cmp) + + if fptrace > 0 { + print("*** cmp D[", regd, "]::0 ", hex(m.fflag), "\n") + } + return 1 + + case 0xeeb50ac0: // F[regd] :: 0 (CMPF) + cmp, nan := fcmp64(f32to64(m.freglo[regd]), 0) + m.fflag = fstatus(nan, cmp) + + if fptrace > 0 { + print("*** cmp F[", regd, "]::0 ", hex(m.fflag), "\n") + } + return 1 + case 0xeeb70ac0: // D[regd] = F[regm] (MOVFD) fputd(regd, f32to64(m.freglo[regm])) |